Telefonaktiebolaget Lm Ericsson (publ)

Szilveszter Nádas

Ferenc Fejes

Gergo Gombos

Sandor Laki

Sándor Laki

Gergő Gombos

Methods and apparatus for packet marking for a Hierarchical Quality of Service (HQoS) is provided to control resource sharing among a plurality of network flows with differentiated services. A packet marker at a single point (e.g., gateway) encodes the resource sharing policy for a plurality of packet flows into a single packet value. The HQoS policy is then realized by using simple PPV schedulers at the bottlenecks in the network.

Marker graph for HQoS

A computing device () identifies a plurality of regions of a Throughput Value Function, TVF (). The TVF () maps throughput values to packet values for a plurality of subflows. The computing device () periodically updates a plurality of matrices comprising a matrix of resource sharing weights and a region determination matrix. The computing device () marks a packet () with a packet value based on the plurality of matrices. The packet () is received in one of the subflows.

Fair hierarchical quality of service marking

A computing device () receives a packet comprised in a flow, and uses a region index table () to identify a first region of a Throughput-Value Function, TVF (), based on a flow rate of the flow. The first region is one of a plurality of sequentially ordered regions across an input domain of the TVF (). The computing device () randomly selects one of a plurality of preconfigured offset values to identify a second region of the TVF () that is lower than the first region of the TVF () by the randomly selected offset value. The computing device () determines a packet value using a TVF table () that maps each of the regions to a respective one of a plurality of packet values. The packet value is mapped to the second region. The computing device () marks the packet with the packet value mapped to the second region.

Packet processing in a computing device using multiple tables

An intermediate device () of a communication network () assigns each of a plurality of received packets () to either a Low Latency, Low Loss, Scalable throughput (L4S) queue () or a non-L4S queue (). Each queue () is associated with one packet value size counter for each of the plurality of packet values (). The intermediate device () increments, for each packet () and by a size of the packet (), the packet value size counter associated with the queue () to which the packet () is assigned and calculates, for each queue (), a congestion threshold value based on the packet value size counters associated with the queue (). The intermediate device () marks one or more of the packets () as having experienced congestion based on the congestion threshold values of the queues ().

Growing community of inventors

Santa Clara, CA, United States of America

Szilveszter Nádas